Association between proteinuria and retinal vein occlusion in individuals with preserved renal function: a retrospective cohort study

Purpose To investigate the association between proteinuria and the incidence of retinal vein occlusion (RVO) in individuals with preserved renal function. Methods We conducted a retrospective cohort study from 2005 to 2020 using the JMDC Claims Database, a large database of health check-ups and administrative claims in Japan. Individuals who underwent annual health check-ups were included. Participants were classified into three groups based on the protein concentration in their urine: negative (<= 10 mg/dl), trace (10-30 mg/dl) and positive (>= 30 mg/dl). We estimated hazard ratios (HRs) for the development of RVO using Cox regression analyses with adjustment for baseline characteristics by the matching weights calculated from the multiple propensity scores. The cumulative incidence of RVO between the weighted groups was calculated using the Kaplan-Meier estimator. Results There were 1 635 212 eligible participants, among whom 2360 developed RVO. The mean follow-up period was 1036 +/- 951 days. The adjusted HRs for RVO were 1.10 (95% confidence interval 0.92-1.33) and 1.46 (1.19-1.78) in the groups with trace and positive proteinuria compared with negative proteinuria respectively. The cumulative incidence rates of RVO at 7 years were 0.40%, 0.43% and 0.55% in the groups with negative, trace and positive proteinuria, respectively. Conclusion Positive proteinuria was independently associated with an increased incidence of RVO in individuals with preserved renal function. Proteinuria may be a novel risk factor for RVO development.